Revised Abstract: Dendritic cells (DC) induce and sustain immune responses and are therefore excellent candidates for active immunization. DC comprise several subsets which may elicit different immune responses in vivo. Our in vitro studies show that both Langerhans cells (LC) and interstitial DC (intDC) are capable of inducing T cell proliferation. However, intDC are uniquely able to induce the differentiation of naive B cells. Though LC express unique properties, such as the presence of Birbeck granules, in vitro studies have not yet attributed a specialized biological function to LC. It is our hypotheses that LC are more efficient than intDC for stimulation of CD8 T cell responses. Because ex vivo cultures of CD34+ hematopoietic progenitors (HPC) yield both LC and intDC, while cultures of monocytes yield interstitial-like DC, we surmise that the presence of both DC subsets allows CD34+HPC-derived DC (cd34DC) to induce stronger immune responses than monocyte-derived DC (MDC). We will study the ability of different DC populations, loaded with melanoma antigens, to elicit melanoma-specific CD4 and CD8 T cell responses in stage IV melanoma. First, we will compare the in vitro immunogenicity of CD34DC and MDC loaded with either melanoma peptides or dying melanoma cells. Second, we will study melanoma-specific immune responses induced in vivo in a clinical trial where two groups of stage IV patients are randomized to receive either melanoma peptides loaded CD34DCs or melanoma peptides loaded MDC. This will allow us to study spreading of the responses beyond the peptides used for the immunization (epitope spreading). Our 4 aims can be summarized as follows: Aim 1: To determine the optimal conditions for the generation of mature CD34DC. Aim 2: To determine whether mature CD34DC induce stronger immune responses in vitro than mature MDC. Aim 3: To determine whether mature CD34DC induce immune responses in vivo than mature MDC. Aim 4: To determine epitope spreading and diversification of melanoma immunity in vitro.